As part of its program to evaluate new treatments, the section is studying the clinical pharmacology of drug of abuse and potential treatment medications. In an already completed study, we showed that buprenorphine, a partial mu-opioid agonist in development for agonist maintenance therapy, has a ceiling for subjective and cardio-respiratory effects and a high safety margin when administered by the IV route. Thus, even if some buprenorphine intended for sublingual administration is illicitly diverted and abused, its margin of safety is likely to be high. More recently, we completed a series of evaluations of cyclazocine, which is a partial agonist at kappa-opioid and mu-opioid receptors. Kappa agonists attenuate a number of cocaine's effects in preclinical studies and, thus, may have potential for the treatment of cocaine dependence. As a first step to test this concept in humans, cyclazocine was evaluated for kappa and mu effects and as a pretreatment to test for attenuation of cocaine effects. Experienced opiate/cocaine users (N = 13) enrolled: 4 dropped out and 9 completed phase 1; 8 completed phase 2. In phase 1, placebo, cyclazocine (0.2, 0.4, 0.8 mg) and hydromorphone (5, 15 mg) were given PO in randomized order in 6 test sessions. In phase 2, cocaine (100 mg intranasal) was given 2 hrs after cyclazocine (0, 0.1, 0.2, 0.4, 0.8, 0 mg, PO, in that order) in 6 sessions conducted daily Monday-Friday and the next Monday. Measures were collected before and after drug administration. Hydromorphone produced prototypic mu-like effects. Cyclazocine had modest kappa-like subjective effects, impaired performance, did not increase urine output, and increased sensitivity to experimental pain. All 4 dropouts in phase 1 occurred after the cyclazocine 0.8 mg dose; this suggests unpleasant effects, a feature of kappa agonists. Consistent with preclinical findings, cocaine effects were consistently lower on the last administration (cyclazocine 0 mg) following 4 days of cyclazocine pretreatment compared to its first administration (cyclazocine 0 mg), though effects of cocaine were not otherwise attenuated. The results of the study provide modest support for the utility of cyclazocine in cocaine treatment, but suggest that kappa agonists could present patient acceptability problems. We are analyzing plasma samples from the study in order to examine the pharmacokinetics and the endocrine effects of the study drugs. In a related study, we evaluated the effects of kappa-opioid agonists on the topography and subjective enjoyment of tobacco smoking; results did not suggest that kappa agonists would be effective for smoking cessation.